The Headless Neural OS for Visual Automation
IRIS GO is an enterprise-grade, local AI orchestration engine. It bypasses standard cloud limitations by running a headless Express + Vite middleware monolith directly on your machine, granting autonomous agent squads root-level access to your terminal, file system, and Git workflows.
[ Watch the Visual Automation Engine in Action ] (Insert a high-quality GIF of your terminal chat UI working here)
Agentic AI CLI + SDK for Real-World Automation
Build, run, and orchestrate autonomous AI agents — locally.
IRIS GO is a developer-first agentic AI platform that enables you to create, execute, and manage autonomous agents directly from your terminal.
It combines:
- ⚡ A powerful CLI (chat + command hybrid)
- 🧠 A local agent runtime (Express + LangGraph)
- 🧩 A flexible SDK for building custom agents
- 🖥️ A pure Vantablack/Neon Emerald Visual Automation Dashboard
IRIS GO is designed for execution-first AI — not just conversation.
You don’t ask. You delegate.
- Autonomous agents that plan → execute → iterate
- Multi-step reasoning using LangGraph
- Real-world task automation
- Conversational + command-based interface
- Run tasks instantly from the terminal
iris "analyze this repo and fix the typescript errors"- File system access
- Terminal execution
- Browser automation
- Custom SDK tools
- Runs on your machine
- No cloud dependency
- JSON / programmatic automation
iris-go/
├── src/
│ ├── surface/
│ ├── kernel/
│ └── cli/
iris start
iris execute "deploy app"
iris chat
import { createAgent } from '@iris/go';
const agent = createAgent({
goal: 'optimize code',
tools: ['fs', 'terminal'],
});
await agent.run();docker build -t iris-go .
docker-compose up -d
Execution > Conversation
Local-first
Developer-first
Use responsibly. System-level execution enabled.
Harsh Pandey
MIT